tabzilla
tabzilla copied to clipboard
Extract remaining metafeature groups
Initially for metafeature extraction, we left out these groups: clustering, complexity, itemset, and concept, which resulted in out of memory errors. We need to investigate whether we can use a cloud instance with sufficient resources to extract any of these metafeatures, or if it is possible to extract them by changing the settings of the metafeature extractor tabzilla_featurizer.py.